The backbone of a bidentate ligand can be described as a scaffold which places the two donor atoms spatially in such a way that they can coordinate efficiently to a metal. Factors such as electron density, provided via the backbone to the donor atoms, or the donor atom-metal-donor atom angle (‘bite angle’) exert a profound effect on the efficacy and specificity of the formed catalyst. Thus, a backbone carrying any combination of the donor atoms or moieties oxygen (hydroxy), nitrogen (amine, amide), phosphorus, arsen or sulfur, forms a ligand. One backbone of particular importance is 1 which is present in the ligands of the DuPHOS-2, PennPhos-3 or BasPhos-type 4. All these ligands are synthesized from phosphine 5 which comprises backbone 1.

The synthesis of precursors for backbone 1, such as 5–8, is difficult because most methods to form a bond between an aromatic carbon and a phosphorus atom fail once there is another substituent present in the ortho-position.
For instance, all known syntheses for ortho-diphosphonates 5 suffer from problems. It has been reported that the Ni-catalysed Arbusov reaction between an aryl bromide and an alkyl phosphite gives a yield of only 14% for an ortho-diphosphonate, whereas for meta- and para-diphosphonates excellent yields of about 90% have been reported (P. Tavs, Chem. Ber. 103, 2428 (1970)). Other approaches employ Diels-Alder chemistry, and here not only the starting materials are difficult to obtain but sometimes additional steps such as aromatisation of the primarily formed Diels-Alder adduct are required. For examples see (i) D. Seyferth and J. D. H. Paetsch, J. Org. Chem. 34, 1969 (1969) (ii) E. P. Kyba et al., Tetrahedron Lett. 22, 1875 (1981); (iii) C. E. Griffin and W. M. Daniewski, J. Org. Chem. 35, 1691 (1970).
Even the optimised route via the photo-initiated Arbuzow reaction between 1,2-dichlorobenzene and trimethyl phosphite requires a reaction period of 5 days and frequent cleaning of the immersion well which renders this route impractical on industrial scale (see E. P. Kyba et al., Organometallics 2, 1877 (1983)).
Once the ortho-diphosphonates 5 have been obtained, their reduction to the synthetically more useful ortho-aryl-bis(phosphine) 8 is difficult (see Organometallics 2, 1877 (1983)), and their direct conversion into the ortho-aryl-bis(dichlorophosphanes) 7 is not possible.
An efficient route to ortho-bis(dialkylamino-phosphines) 6 would be far more useful, as treating compounds such as 6 with HCl readily gives the derivative 7 which can be readily reduced to the primary phosphines 8.
One route for preparing compounds such as 6 involves the reaction of 1,2-dilithiobenzene with chloro-bis(dimethylamino)phosphine. The major disadvantage, however, is here the use of 1,2-di-(mercurio)-benzene which is highly toxic (see K. Drewelies, H.-P. Latscha, Angew. Chem. 94, 642 (1982)). Another attempt to prepare 6 by the reaction of the lithiated diaminophospine borane complex with 1,2-diodobenzene gave selectively only the monosubstituted iodoarylphosphine in 66% yield (see A. Longeau and P. Knochel, Tetrahedron Lett. 37, 6099 (1996).
Flexible manufacturing procedures that allow to attach two donor atoms in the cis-position of a double bond are of value as they allow to increase the range of useful backbones and the derived ligands. Of particular usefulness in the synthesis of bidentate ligands are bis(substituted, e.g. alkyl)-phosphines such as 6, because they can also readily be converted via the bis(dihalogenophosphine) compounds into bisphosphines, both of which are useful as intermediates for other ligands.
A goal of the present invention is to allow for an easy and simple manufacturing process for many ligands, said manufacturing process being economically and technically advantageous, implying simple steps, controlled costs and industrial scale applicapility. A further goal is to provide novel ligands that allow to obtain stable complexes that are useful especially as chiral catalysts with good reactivity and allowing for reactions with high regio-, chemodiastereo- and/or enantioselectivity, thus making it possible to use them e.g. for stereocontrolled reactions, advantageously under mild reaction conditions while maintaining appropriate reaction rates. A goal is also to establish new ligands capable of forming metal complexes that permit stereocontrolled reactions, particularly reduction (especially hydrogenation) or isomerisation reactions, and the appropriate reaction conditions that facilitate the production of optically active products with high diastereomeric or enantiomeric excess.